Turning an institution’s unstructured record into a structured, searchable corpus it owns outright. Grant-funded, delivered against a written scope, reported honestly.
What is at risk
Most long-lived organisations are carrying an archive they cannot use. Three decades of a mailing list that exists only as mail exports. A newsletter run held as scanned pages with no text layer. A previous website in a format no current browser renders, whose links point at a server that was decommissioned years ago. Photographs in a box. Minutes in a filing cabinet. A founder’s papers on a drive nobody has mounted since the founder died.
It is not lost, exactly. It is simply unreachable, and every year makes it more so.
What we do with it
- Recover. Extract from dead formats, obsolete media, mail archives and static pages, including files whose original structure has to be reconstructed.
- Process. Optical character recognition, transcription, translation where the collection is not in English, and extraction of names, dates, places and topics. Machine-assisted, at a scale that was not affordable five years ago.
- Structure. Catalogue into a real schema with consistent metadata, so the collection can be searched, filtered and cited rather than merely browsed.
- Publish. Under the institution’s own name, on infrastructure it owns, with permanent URLs that will still resolve in a decade.
- Verify. Counts reconciled at every stage — items extracted against items published, images referenced against images present — because an archive with silent gaps is worse than no archive.
How it is paid for
Rarely from an operating budget. This work is usually grant-funded, and the grant goes to the institution, not to us. We are the delivery vendor and, where it helps, we assist with the technical and budget sections of the application.
That arrangement puts a specific obligation on us. Funders in this field have been burned by projects that drew down money and produced little, with no accountability for the result. So every engagement carries a defined scope, dated work phases, a line-item budget, and a written final report showing what was delivered against what was promised — the documents a funder needs in order to be willing to fund the next one.
A note on faithfulness
We publish what the institution published. We do not silently correct, reinterpret or filter historical material to suit present sensibilities, and we do not substitute our judgement for the institution’s on what its own record contains. Where material raises a genuine question — personal data, a living dispute, a restriction on use — we bring it to the institution rather than deciding alone.